A strict communication protocol designed to orchestrate multiple AI agents within a single Telegram group chat while preventing infinite loops and chaotic interactions.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install telegram-agent-coordination
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install telegram-agent-coordination using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Telegram Agent Coordination?
The Telegram Agent Coordination skill provides a robust framework for managing complex interactions between multiple AI agents and human operators in a shared Telegram environment. By utilizing Openclaw Skills, developers can implement a disciplined hierarchy consisting of a Human Founder, a COO bot, and Worker bots. This structure ensures that every message serves a specific purpose and that agents only interact when valid triggers are met, effectively eliminating the risk of recursive bot-to-bot chat loops.
This skill is essential for scaling AI operations where specialized agents must collaborate on a single mission. It replaces unstructured chatter with a rigorous lifecycle of task assignment, status reporting, and mission completion. By enforcing sender validation and visible message prefixes, the protocol maintains a clean, searchable, and manageable command-and-control history within the Telegram interface.
Telegram Agent Coordination Use Cases
- Orchestrating a COO bot and multiple execution bots in a single team chat.
- Preventing infinite reply loops when two or more LLM-based agents share a channel.
- Managing complex multi-step missions that require human oversight and approval.
- Creating a standardized command-and-control interface for distributed AI teams.
- Escalating technical blockers from worker agents to human supervisors through a managerial bot.
How Telegram Agent Coordination Works
- The Human Founder initiates a mission by sending a START MISSION command to the group.
- The COO bot validates the human sender, acknowledges the mission, and issues a TASK to the relevant Worker bot.
- The Worker bot validates that the message came from the COO, sends an ACK (acknowledgment), and begins execution.
- During execution, the Worker may provide STATUS updates or flag a BLOCKER if external input is needed.
- Upon completion, the Worker sends a RESULT message back to the COO.
- The COO bot reviews the result, summarizes the mission achievements, and issues a MISSION COMPLETE message to the human.
Telegram Agent Coordination Setup
To implement this coordination protocol using Openclaw Skills, ensure your bots have appropriate group permissions and follow these steps:
# Define your agent roles in your configuration
# Ensure all bots have the protocol skill enabled
# Configure sender allowlists for the COO and Worker bots
# Example: Initializing the group environment
/set_role bot_1 commander
/set_role bot_2 executor
/allowlist_sender human_id
Ensure that the MISSION START:, TASK:, and RESULT: prefixes are hardcoded into your agent prompt templates.
Telegram Agent Coordination Data Schema & Taxonomy
The protocol organizes communication using a structured prefix system and role-based validation logic:
| Message Component | Description | Examples |
|---|---|---|
| Visible Prefix | Mandatory header identifying message type. | MISSION START:, TASK:, BLOCKER: |
| Sender ID | Unique Telegram ID used for validation. | founder_01, coo_bot_02 |
| Role Mapping | Logic that maps IDs to specific permissions. | Founder, COO, Worker |
| Message Types | Restricted set of allowed intents. | STATUS, RESULT, ACK, REWORK |
Telegram Agent Coordination Advanced Features
- Anti-Loop Logic: Automated silence triggers if an agent detects its own message as the last in a thread.
- Blocker Escalation: Structured
NEEDED:syntax to pinpoint missing resources for rapid human intervention. - Mission Lifecycle Control: Capabilities for priority changes and mission stops by the Human Founder.
- Role-Based Persona Enforcement: Managerial style for COOs and execution-focused style for Workers.
- Parallel Workstreams: Support for the COO to manage multiple tasks across different workers simultaneously.
SKILL.md
Loading
Related Openclaw Skills
E-Commerce Product Scene Generator
A specialized AI agent skill designed to transform product descriptions into high-end, physics-grounded image generation prompts for professional e-commerce visuals.
Jira Expert
A master-level AI agent skill for technical Jira operations, including complex JQL queries, automated workflow design, and comprehensive project configuration.
COO Commander
A strategic orchestration skill that acts as a mission controller to manage, delegate, and track tasks across worker agents.
Feishu Image Messaging
A powerful automation tool for sending local or remote images and mixed text-image messages via the Feishu API.
Indigo IPFS & Collector
A decentralized storage and blockchain utility skill for interacting with IPFS and Indigo Protocol collector UTXOs.
Feishu Group Memory
A powerful AI skill that transforms chaotic Feishu group conversations into structured, searchable records and actionable business intelligence.
Featured*
